Direct API deployment
A Direct API deployment exposes your assistant over authenticated REST endpoints — for mobile apps, backend services, and custom UIs that do not use the web widget.
When to use Direct API
Section titled “When to use Direct API”| Direct API | Web widget |
|---|---|
| Native apps, server-side chat | Embedded website chat |
| You control the entire UI | Chatevo provides chat UI + SSE streaming |
| Bearer/API key from your backend | Public embed with domain allowlist |
Create deployment
Section titled “Create deployment”- Deployments → Create → Direct API.
- Name and Assistant selection.
- Environment — development, staging, production.
- Note deployment ID and create API keys for your server.
- Configure rate limits as needed.
Send messages
Section titled “Send messages”Use the Messages API with your deployment context:
- Create or continue a conversation.
- POST a message — assistant runs RAG and tools same as widget.
- Optional: consume streaming responses if your client supports SSE.
Reference: API overview, OpenAPI spec.
Tools and rich content
Section titled “Tools and rich content”Tool execution is identical to the widget. API responses may include structured rich content payloads when result_transform applies — your client renders them or maps to your UI.
Security
Section titled “Security”- Never expose API keys in mobile apps — proxy through your backend.
- Rotate keys on compromise; per-environment keys recommended.
- Same credential vault powers tools — not exposed via Direct API.